There is considerable evidence that a significant proportion of patients fail to benefit from psychotherapy. Thus it is critical to identify those patients who are at risk for treatment dropout or poor outcome, and to develop ways of improving the likelihood that they will complete treatment and benefit from it. There is also evidence that a significant proportion of patients present with comorbid conditions and that those with personality disorders are most treatment resistant and challenging to the therapeutic alliance. In addition, there are findings suggesting that recognizing and addressing ruptures in the therapeutic alliance may be important to treatment completion and good outcome. We propose that the therapeutic alliance is a useful means for adapting treatment adherence to the context of psychotherapy. Finally, the research literature demonstrates the relevance of individual differences to treatment outcome, including resolving ruptures in the therapeutic alliance. In light of this background, we argue for the importance of conducting research on ruptures in the therapeutic alliance and on training therapists to work effectively in resolving them, as well as research designs that attempt to control for individual differences. The primary aim is to conduct a pilot study involving the implementation of an innovative design to evaluate the usefulness of a specialized training for therapists in the management of the therapeutic alliance. Specifically, this aim includes: (a) To demonstrate the feasibility of a study protocol that will include the implementation of a multiple phase change (or multiple baseline) design in which therapists trained to conduct a cognitive-behavioral therapy for personality disorders are introduced at different intervals to additional training, specifically designed to improve their skills in the recognition and resolution of alliance ruptures; and (b) To collect preliminary data on the additive effect of this specialized training on the treatment process (i.e., the interpersonal interactions between therapist and patient). The secondary aim of our application is to conduct a psychometric study that develops three strategies to assess ruptures in the therapeutic alliance and their resolution. The strategies include (a) patient-report and (b) therapist-report measures rated in a post-session questionnaire and (c) an observer-rated measure designed to capture specific patient behaviors and therapist interventions. The study aims to demonstrate the psychometric properties of these measures, by conducting item analyses, tests of internal consistency and interrater reliability (where appropriate). The study will also establish the construct validity of these measures, by examining convergent and discriminant relations among various measures of session impact and treatment outcome. These specific aims are essential first steps towards our long-term objective, which is to conduct a large-scale study that demonstrates the impact of the specialized training in rupture recognition and resolution on treatment adherence and outcome, as well as explores a variety of factors related to the interaction of therapist individual differences with training procedures. [unreadable] [unreadable]